Novel Histamine H3-Receptor Antagonists with ... - ACS Publications

Sep 21, 2000 - Novel Histamine H3-Receptor Antagonists with Carbonyl-Substituted 4-(3-(Phenoxy)propyl)-1H-imidazole Structures like Ciproxifan and ...
0 downloads 0 Views 95KB Size
J. Med. Chem. 2000, 43, 3987-3994

3987

Novel Histamine H3-Receptor Antagonists with Carbonyl-Substituted 4-(3-(Phenoxy)propyl)-1H-imidazole Structures like Ciproxifan and Related Compounds† Holger Stark,*,# Bassem Sadek,# Michael Krause,# Annette Hu¨ls,# Xavier Ligneau,‡ C. Robin Ganellin,| Jean-Michel Arrang,§ Jean-Charles Schwartz,§ and Walter Schunack# Institut fu¨ r Pharmazie, Freie Universita¨ t Berlin, Ko¨ nigin-Luise-Strasse 2+4, 14195 Berlin, Germany; Laboratoire Bioprojet, 30 rue des Francs-Bourgeois, 75003 Paris, France; Department of Chemistry, Christopher Ingold Laboratories, University College London, 20 Gordon Street, London WC1H 0AJ, U.K.; and Unite´ de Neurobiologie et Pharmacologie Mole´ culaire (U. 109), Centre Paul Broca de l’INSERM, 2ter rue d’Ale´ sia, 75014 Paris, France Received April 6, 2000

Novel histamine H3-receptor antagonists possessing a 4-(3-(phenoxy)propyl)-1H-imidazole structure generally substituted in the para-position of the phenyl ring have been synthesized according to Mitsunobu or SNAr reactions. With in vitro and in vivo screening for H3-receptor antagonist potency, the carbonyl-substituted derivatives proved to be highly active compounds. A number of compounds showed in vitro affinities in the subnanomolar concentration range, and the 4-hexanoyl (10) and 4-acetyl-3-methyl (29) substituted derivatives showed in vivo antagonist potencies of about 0.1 mg/kg after po administration. Many proxifans were also tested for their affinities at other histamine receptor subtypes thereby demonstrating their pronounced H3-receptor subtype selectivity. Since the cyclopropyl ketone derivative 14 (ciproxifan) had high affinity in vitro as well as high potency in vivo, it was selected for further studies in monkeys. It showed good oral absorption and long-lasting, dose-dependent plasma levels making it a promising compound for drug development. Introduction Histamine is a well-known messenger in various (patho)physiological conditions. Receptor-dependent effects are mediated by three different established proteins: H1, H2, and H3 receptors.3,4 Whereas the H1 and H2 receptors are located postsynaptically, the H3 receptors can be located both pre- and postsynaptically. Discovery of the H3 receptor helped to clarify the neurotransmitter function of histamine. On H3-receptor activation, inhibition of histamine release5 and inhibition of histamine synthesis6 have been observed on histaminergic axon terminals. H3-Receptor antagonists increase histaminergic neuron activity by inhibition of this feedback mechanism. H3 Heteroreceptors on nonhistaminergic neurons in the brain as well as in the periphery influence a portfolio of different classical neurotransmitters and neuropeptides.7 Therefore, numerous potential therapeutic targets for histamine H3receptor antagonists have been proposed, most of them derived from pharmacological in vitro or in vivo studies. Since the highest density of H3 receptors is found in distinct areas of the central nervous system (CNS), the treatment of neurological and psychiatric diseases seems to be most promising, e.g., schizophrenia, attention-deficit hyperactivity disorder (ADHD), dementia, epilepsy, or narcolepsy.8 † Presented in part at the XVth International Symposium on Medicinal Chemistry, Edinburgh, U.K., Sept 6, 1998,1 and WO Patent 96/29 315.2 * To whom correspondence should be addressed. Tel: +49 30 838 53928. Fax: +49 30 838 53854. E-mail: [email protected]. # Freie Universita ¨ t Berlin. ‡ Laboratoire Bioprojet. | University College London. § Centre Paul Broca de l’INSERM.

Chart 1. Histamine H3-Receptor Antagonists

Developments in the medicinal chemistry field of histamine H3-receptor antagonists have yielded numerous compounds that are highly active in vitro and some of them also in vivo. Although these compounds possess different functionalities, they all have more or less in common an imidazole heterocycle connected by a spacer with a polar group being optionally connected by another spacer to a lipophilic group.8 With this modular construction pattern it seems that one element can be replaced by another structurally different moiety if the rest of the molecule displays enough receptor affinity. Classical antagonists are thioperamide and clobenpropit (Chart 1). Newer antagonists such as UCL 1390,9 FUB 470,10 or the (un)labeled iodoproxyfan,11 which displays partial agonism in some test models, have been described which offer some potential benefit in affinity, selectivity, and/or toxicity. Most recently the chiral compound GT-233112 (Perceptin) has been described as

10.1021/jm000966l CCC: $19.00 © 2000 American Chemical Society Published on Web 09/21/2000

3988

Journal of Medicinal Chemistry, 2000, Vol. 43, No. 21

Stark et al.

Scheme 1. General Procedures for Synthesis of Arylproxifansa

a Trt, trityl; (a) Ph P, DEAD, corresponding phenol, THF; (b) HCl, acetone; (c) (i) NaH, toluene, (ii) corresponding fluoroaromate, (iii) 3 HCl, THF.

entering phase II in clinical trials for the treatment of ADHD.13 The outcome of this first clinical trial of H3receptor antagonists and the recent cloning of human14 and rodent15 histamine H3 receptors will highly influence further developments. Since compounds such as UCL 1390 and FUB 470, which have an imidazolylpropoxy moiety in common, have high in vivo activity (cf. Table 1) we selected the para-substituted 4-(3-(phenoxy)propyl)-1H-imidazole as a pharmacophore for further elaboration. The triple bond of the cyano or alkynyl group should be replaced by other moieties, preferably by electron-withdrawing groups which are expected to be nontoxic. This study is focused on carbonyl-substituted derivatives since the antagonist potency of these compounds was found to be higher than that of compounds with other substituents. The influence of the acyl group in the para-position of the phenyl ring, the influence of the position of the carbonyl group in the para-substituents, and the effects of additional substituents in the phenyl ring were investigated. Synthesis and pharmacological H3-receptor antagonist screening of the compounds are described to provide information for structure-activity relationships (SARs). Additional pharmacological screening for histamine receptor subtype selectivity, histamine-Nmethyltransferase inhibitory activity, and in vivo kinetics was also performed for selected compounds. Chemistry The key intermediate for all the compounds was the trityl-protected 3-(1H-imidazol-4-yl)propanol which was conveniently prepared from commercially available urocanic acid.11a Ether formation was effected by a Mitsunobu type synthesis16 with the corresponding phenol derivative as described before (Scheme 1).9 Deprotection under acidic conditions resulted in the final products (1-4, 6-12, 14-18, 20, 22-24, 27-31, 38, 39, and with modifications for 13 and 25). The mild conditions using the Mitsunobu protocol were useful for almost all carbonyl-substituted phenol derivatives which tend to undergo unwanted side reactions under different reaction conditions. Alternative methods such as the Williamson synthesis with 4-(3-chloropropyl)-1H-imidazole17 were not successful due to intramolecular cyclization and partial polymerization under basic conditions. The corresponding carbonyl-substituted phenol derivatives were prepared by Friedel-Crafts alkylation of phenol in nitrobenzene at ambient or slightly increased temperature using an excess of AlCl3. Mixtures of isomeric ortho- and para-substituted phenol derivatives when

Scheme 2. Synthesis of 13a

a Trt, trityl; (a) EtOH, HCl, acetone, reflux, 45 min or HCOOH (90%), ambient temperature, 24 h.

obtained were readily separated by standard chromatographic procedures. The phenol for the synthesis of 24 was prepared by aldol reaction with 4-hydroxybenzaldehyde and acetone under standard conditions. In cases where the corresponding methoxy derivative was commercially available (14, 36, 37) the requisite phenol was obtained most conveniently by ether cleavage using BBr3‚etherate in dichloromethane at low temperature according to known methods.10 Another method for ether formation is a nucleophilic substitution (SNAr) on fluoroaromatics containing additional electron-withdrawing substituents. This reaction was performed with the sodium salt of the synthon as nucleophile and corresponding fluorophenyl derivatives, followed by deprotection for compounds 19, 26, and 32-35 (Scheme 1). The yields vary to a large extent depending on the aryl substituents, but this is an expedient way to obtain the final compounds from commercially available starting materials. The benzyl alcohol 5 was obtained by reduction of 4 with complex hydrides (e.g., LiAlH4). The carboxylic acid 21 was prepared by acidic ester hydrolysis of 20. As shown in Scheme 2 compound 13 was obtained from the corresponding alkynyl-substituted phenyl ether which was prepared by Heck reaction of (4-iodophenoxy)methoxymethane (Mom-protected 4-iodophenol) with 3,3-dimethylbutyne, phenol deprotection with sodium ethanthiolate, and following Mitsunobu reaction with the key synthon, 3-(1-trityl-1H-imidazol-4-yl)propanol. The tritylated alkynyl precursor molecule obtained was prepared according to a method described before for related tert-butoxycarbonyl (Boc)-protected imidazole derivatives for the synthesis of alkynylphenyl derivatives.10 Deprotection under acidic conditions led to an addition of water to the C-C triple bond and formation of the tautomeric carbonyl isomer 13 (Scheme 2). Taking corresponding trimethylsilyl- or alkyl-substituted alkynyl precursor molecules for the Heck reaction and performing the analogous reaction steps, compounds 6 or 7 and 8, respectively, were also prepared by this

Novel Histamine H3-Receptor Antagonists

Journal of Medicinal Chemistry, 2000, Vol. 43, No. 21 3989

Scheme 3. Synthesis of 25a

affinity compared to the rather low number of other substituents previously tested (F, CF3, CtC-R with R ) CH3, C3H7, C(CH3)3, Si(CH3)3)9,10 and in this series. Moreover, steric factors seem to be more important than electronic properties. Larger alkyl substituents on the carbonyl moiety slightly decreased affinity (e.g., 6, 7 f 8 f 9, 10 f 11; 12 f 13; 14 f 15 f 16). Compounds possessing moieties with additional binding properties distant to the carbonyl group, e.g., 18-21, showed obviously decreased affinity. The steric factor is also important within the series of polysubstituted derivatives (Table 2). Two methyl groups in the neighborhood (ortho-related) of the acetyl group strongly decreased affinity, whereas the second methyl group in another position (meta-related) exhibited a less pronounced effect (30 f 31). Affinity was also decreased when other groups adjacent to the acetyl group could influence its orientation (32, 33). In this series small lipophilic substituents on the carbonyl group seem to be optimal since the acetyl (6), propionyl (7), cyclopropylcarbonyl (14), and cyclobutylcarbonyl (15) derivatives displayed subnanomolar affinities. In Vivo Screening in Mice CNS. The novel compounds were also screened for their effect at histamine H3 receptors in vivo on mice brain cortex after po administration measuring the increase in Nτ-methylhistamine level, the main histamine metabolite.19 All those compounds where potency could be detected in vivo had antagonist properties at H3 receptors. Surprisingly, the simple phenoxy derivative 1 as well as the derivatives with hydroxy groups (2, 5, 21) did not show any in vivo potency. Compared to their relative high in vitro affinity, pharmacokinetic properties seem to be one of the most probable reasons for the in vivo failure of these compounds as well as of 4, 18-25, 27, 30, and 33. Our present data do not show direct correlation between the potencies of compounds as antagonists in vitro and in triggering histamine neuron activation in vivo. Very recently it has also been shown that not all histamine H3-receptor antagonists influence the Nτmethylhistamine CNS level to the same extent although they are reaching brain areas.20 Therefore, it can be concluded that despite the direct antagonist effect on H3 receptors other factors that influence the Nτ-methylhistamine level cannot be excluded. This complex topic needs further investigation because different G protein coupling, different pharmacology of somatodendritic vs presynaptic H3 receptors, receptor cross-reactivity, influence on enzymes of histamine metabolism, etc., may be reasons for these apparent discrepancies. The statement concerning the electronic and steric effect for the in vitro affinity also applies for the in vivo potency. Different “proxifans”, which are compounds possessing a 3-(1H-imidazol-4-yl)propoxy moiety, showed strikingly high in vivo potency. Antagonists 3, 6-11, 13-17, 28, 29, 31, 32, and 34 were active in a dose range below 1 mg/kg after peroral administration and compounds 10 and 29 even below 0.1 mg/kg. For the different n-alkyl chains on the carbonyl group (6-21) increasing chain length slightly increased potency reaching an optimum with the hexanoyl moiety (10). Its isomeric compound 13 was significantly less effective suggesting that in addition to the presumed benefit of

a Trt, trityl; (a) HCtCMgBr, THF, reflux, 1 h; (b) EtOH, HCl, acetone, reflux, 45 min.

method, but total yields were lower as compared to the Mitsunobu protocol mentioned before (results not shown). Another indirect way of preparation was applied for compound 25. The tritylated precursor of the benzaldehyde compound 4 was isolated and subjected to a Grignard reaction with ethynylmagnesium bromide. This propargyl alcohol intermediate rearranged into the cinnamaldehyde derivative 25 in the course of acidcatalyzed detritylation (Scheme 3). This isomerization is known as Meyer-Schuster rearrangement and describes the acid catalyzed 1,3-rearrangement of secondary or tertiary alcohols with an ethynyl functionality leading to R,β-unsaturated carbonyl derivatives via the corresponding ketenol intermediate (Scheme 3).18 By this rearrangement terminal alkynes result in corresponding aldehydes as shown with 25. It should also be possible to get this compound directly by reacting 4-hydroxycinnamaldehyde with the key intermediate followed by deprotection. But the tritylated propargyl intermediate was available to us from previous studies for the design of histamine H3-receptor antagonists,10 and therefore, this was the most convenient route for the preparation of 25. Pharmacological Results and Discussion In Vitro Assay on Synaptosomes of Rat Cerebral Cortex. The novel compounds were tested for their effect at histamine H3 receptors in vitro on synaptosomes of rat cerebral cortex for the release of [3H]histamine (Tables 1 and 2).19 All compounds had antagonist properties at H3 receptors in vitro. Most compounds displayed antagonist potencies in the nanomolar or even subnanomolar concentration range. It can be clearly seen that the electronic properties of the para-substituents do not influence affinity to a large extent because compounds possessing electron-donating substituents (2, 3, 5, 22, 23, 38, 39) showed comparable potency to those with electron-withdrawing substituents (e.g., 4, 6, 24-26, 36, 37). The same is true for polysubstituted derivatives (e.g., 6 f 28 f 29, 31 f 32; Table 2). Nevertheless, it should be stated that most acyl derivatives possess a tendency for higher in vitro

3990

Journal of Medicinal Chemistry, 2000, Vol. 43, No. 21

Stark et al.

Table 1. Structures, Physical Data, and Pharmacological Screening Results of Para-Substituted 4-(3-(Phenoxy)propyl)-1H-imidazoles for Histamine H3-Receptor Antagonist Potency in Vitro and in Vivo in Rodents

no.

R

formula

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 UCL 1390 FUB 470 thioperamide clobenpropit

H OH OCH3 CHO CH2OH CO-CH3 CO-C2H5 CO-C3H7 CO-C4H9 CO-C5H11 CO-C6H13 CO-CH(CH3)2 CO-CH2-C(CH3)3 CO-cyclopropyl CO-cyclobutyl CO-cyclopentyl CO-CH2-Ph CO-(CH2)3-piperidino CO-(CH2)3-pyridoindole CO-(CH2)3-COOC2H5 CO-(CH2)3-COOH CH2-CO-CH3 (CH2)2-CO-CH3 (CH)2-CO-CH3 (CH)2-CHO SO2-CH3 cyclopentyl (R ) C≡N)h (R ) C≡CH) j

C12H14N2O‚0.95C2H2O4 C12H14N2O2‚C4H4O4 C13H16N2O2‚C4H4O4‚0.25H2O C13H14N2O2‚C4H4O4 C13H16N2O2‚C4H4O4‚0.25H2O C14H16N2O2‚C4H4O4 C15H18N2O2‚C4H4O4 C16H20N2O2‚C4H4O4‚H2O C17H22N2O2‚C4H4O4 C18H24N2O2‚C4H4O4 C19H26N2O2‚C4H4O4‚0.25H2O C16H20N2O2‚C4H4O4‚0.5H2O C18H24N2O2‚C4H4O4 C16H18N2O2‚C4H4O4 C17H20N2O2‚C4H4O4‚0.25H2O C18H22N2O2‚C4H4O4‚0.25H2O C20H20N2O2‚C2H2O4‚0.75H2O C21H29N3O2‚2C2H2O4 C30H29FN4O2‚2C4H4O4‚3H2O C19H24N2O4‚C4H4O4‚H2O C17H20N2O4‚HCl‚2H2O C15H18N2O2‚C4H4O4‚0.5H2O C16H20N2O2‚C4H4O4‚0.5H2O C16H18N2O2‚C4H4O4‚0.5H2O C15H16N2O2‚C4H4O4‚0.5H2Og C13H16N2O3S‚C4H4O4 C17H22N2O‚0.75C2H2O4

Mr

yield (%)

mp (°C)

Kia (nM) xj ( sxj

ED50b (mg/kg) xj ( sxj

278.8 334.3 352.9 346.3 352.9 360.4 374.4 406.4 402.4 416.5 435.0 397.4 416.5 386.4 404.9 419.0 423.9 535.5 782.8 478.5f 388.8 383.4 397.4 395.4 381.4 396.4 337.9

35 30 85 85 50 80 80 75 85 75 85 85 70 55 80 80 65 20 10 55 80 75 70 45 35 25 60

163-165 138 120-121 120 134 118 136 87 104-105 113-115 116-118 95 139 118-120 130 141 174-176 190-192 70 oil 97 89 104 123 119-121 135 224-225

55 ( 11 18 ( 4 2.2 ( 0.4 12 ( 3 8.9 ( 1.5 0.80 ( 0.16 0.74 ( 0.07 1.9 ( 0.4 2.6 ( 0.7 2.5 ( 0.3 7.0 ( 1.3 4.8 ( 0.6 27 ( 10 0.49 ( 0.09c 0.89 ( 0.18 7.2 ( 1.1 5.9 ( 2.4 ndd 410 ( 139 122 ( 17 49 ( 10 53 ( 11 91 ( 20 20 ( 5 32 ( 7 40 ( 10 nd 12 ( 3 2.3 ( 0.8 4 ( 1c 0.6 ( 0.1c

>10 >10 0.20 ( 0.08 ∼30 >10 0.24 ( 0.06 0.27 ( 0.09 0.17 ( 0.05 0.15 ( 0.04 0.098 ( 0.029 0.15 ( 0.05 3.9 ( 1.9 0.85 ( 0.14 0.14 ( 0.03c 0.19 ( 0.09 0.85 ( 0.33 0.46 ( 0.11 >10 >10 >10 >10 >10 >10 >10 ∼30 8.8 ( 2.2 ∼30 0.54 ( 0.23 0.12 ( 0.07 1.0 ( 0.5c 26 ( 7c

a Functional H -receptor assay in vitro on synaptosomes of rat cerebral cortex.19 b Central H -receptor screening in vivo after oral 3 3 administration to mice.19 c Ref 21a and literature therein. d nd, not determined. e Pyridoindol, 8-fluoro-2,3,4,5-tetrahydro-1H-pyrido[4,3f g h j b]indol-2-yl. Elemental analyses not performed. C,H; N: calcd, 7.35; found, 8.02. Ref 9. Ref 10.

Table 2. Structures, Physical Data, and Pharmacological Screening Results of Polysubstituted 4-(3-(Phenoxy)propyl)-1H-imidazoles for Histamine H3-Receptor Antagonist Potency in Vitro and in Vivo in Rodents

no.

R1

R2

28 29 30 31 32 33 34 35 36 37 38 39

CO-CH3 F CO-CH3 CH3 CO-CH3 CH3 CO-CH3 H CO-CH3 CF3 CO-CH3 OCH3 CO-C2H5 F CO-cyclopropyl H -CO-CH2-CH2-CO-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-

R3

R4

formula

H H CH3 H H H H H H H H H

H H H CH3 H H H NO2 H H H H

C14H15FN2O2‚C4H4O4‚0.25H2O C15H18N2O2‚C4H4O4 C16H20N2O2‚C4H4O4‚0.5H2O C15H18N2O2‚C4H4O4‚0.25H2O C15H15F3N2O2‚C2H2O4‚0.5H2O C15H18N2O3‚C4H4O4‚0.5H2O C15H17FN2O2‚C4H4O4‚0.25H2O C16H17N3O4‚C4H4O4 C15H16N2O2‚C4H4O4‚H2O C16H18N2O2‚C4H4O4‚0.5H2O C15H18N2O‚0.8C2H2O4 C16H20N2O‚0.85C2H2O4

Mr

yield (%)

mp (°C)

Kia (nM) xj ( sxj

ED50b (mg/kg) xj ( sxj

382.9 374.4 397.4 387.9 411.3 399.4 396.9 431.4 390.4 395.4 314.3 332.9

80 80 85 80 55 60 35 60 40 75 40 20

114 109 92 126 117 124 119-121 151-152 147 102 188-190 171-173

1.1 ( 0.2 1.2 ( 0.3 90 ( 19 4.8 ( 0.8 17 ( 5 325 ( 164 57 ( 11 ndc 17( 4 13 ( 3 38 ( 6 45 ( 14

0.15 ( 0.06 0.08 ( 0.02 >10 0.39 ( 0.10 0.64 ( 0.14 >10 0.9 ( 0.2 5.5 ( 2.8 1.5 ( 0.5 7(4 1.3 ( 0.2 3.3 ( 1.3

a Functional H -receptor assay in vitro on synaptosomes of rat cerebral cortex.19 3 administration to mice.19 c nd, not determined.

increased lipophilicity, steric factors have to be considered. The steric demands of this alkyl substituent were

b

Central H3-receptor screening in vivo after oral

impressively shown by the isopropyl derivative 12 compared to its cyclized analogue 14. In particular, the

Novel Histamine H3-Receptor Antagonists

Journal of Medicinal Chemistry, 2000, Vol. 43, No. 21 3991 Table 3. Activity of Selected Compounds at Histamine Receptor Subtypes

Figure 1. Kinetics of ciproxifan (14) in monkeys (n ) 6-11) after single oral administration.

cyclopropyl derivative 14 is more than 25 times more potent than the isopropyl derivative 12. Even the homologous cycloalkyl derivatives (15, 16) are more potent than 12. Shifting of the carbonyl moiety away from the phenyl ring (22-25), its exchange to sulfone (26), or its omission (27) led to strong decreases in potencies. The carboxylic acid 21 was synthesized for the preparation of antibodies for the development of a specific radioimmunoassay for the main compounds of this series (Ligneau et al., unpublished results). Additional substitution of the phenyl ring was accepted when the substituents did not influence too much the little steric demands of the carbonyl moiety. One methoxy group (33) or two methyl groups (30) in the ortho-position relative to the carbonyl group led to inactive compounds, whereas a nitro group in the corresponding meta-position (35) only caused a decrease. Bicyclic compounds (36-39) gave no improvements. Additional substitution was tolerated with one fluoro (28, 34), methyl (29, 31), or trifluoromethyl (32) group since it led to retention of activity. Although compounds 28 and 29 showed an improvement in potency compared to 6, this cannot be taken as a general rule (cf. 7 f 34). Compounds 10, 14, and 29 were the most active antagonists in vivo in this series and are also among the most active ones known so far to our knowledge. The potencies of these compounds were not statistically significantly different from each other. Compound 14, named ciproxifan, was selected for further studies because it showed extremely high activity in vitro as well as in vivo. With these studies ciproxifan appears to be an orally bioavailable, extremely potent, and selective histamine H3-receptor antagonist whose vigilance- and attention-promoting effects are promising for therapeutic applications in aging or degenerative disorders.21 Oral bioavailability and detectable plasma levels for more than 6 h with different doses of ciproxifan are shown by the preclinical kinetic data in monkeys in Figure 1 although rapid formation of an active metabolite cannot be excluded by the method of determination. The areas under the curves (AUC) were 6.2, 26.6, and 61.6 µM/h for the 3, 10, and 30 mg/kg po doses, respectively. Furthermore, ciproxifan showed low inhibitory effect on the histamine-metabolizing enzyme, histamine-N-methyltransferase (E.C. 2.1.1.8) (IC50 ) 9.0 ( 1.2 µM), indicating that this enzyme is not influenced in therapeutic concentrations. Ciproxifan also showed low inhibitory affinity for the cytochrome P450 enzyme

no.

H3 pKia

H3 pKbb

H2 pKbc

H1 pKbd

3 4 5 6 7 8 9 10 12 14f 15 16 22 23 28 29 32

8.7 7.9 8.1 9.1 9.1 8.7 8.6 8.6 8.3 9.3 9.1 8.1 7.3 7.1 8.9 8.9 7.8

7.4 6.3 6.7 7.8 8.0 8.1 8.4 8.4 7.1 8.4 8.5 7.9 6.2 6.9 7.9 7.9

4.3

4.0

4.1e 4.6 4.9 5.4 4.7 5.3 4.9 5.4 5.8

4.2e 5.6 4.8 5.1 5.3 4.9 4.6